This invention is directed to a surveying technique for determining whether a monitored individual is tuned to a given signal source such as a television channel or radio station and, in particular, to the transmission of a survey signal combined with a programming signal which are both in the audible frequency range, but which converts and reproduces the survey signal to an inaudible signal that is used to detect the signal source to which the individual is tuned.
It is important for a number of reasons to survey an audience to determine to what extent each of its members is tuned at any given time to a particular source of programming such as a television channel or radio station (collectively referred to as a "signal source"). Advertisers are, of course, interested in determining the number of people exposed to their broadcast commercials and to characterize their listeners by economic and social categories. Broadcasters find the statistics regarding audience size and type beneficial in setting their advertising rates.
Prior art techniques for obtaining such information involve primarily the following approaches. People within the range of the radio station or who receive a television channel (either over the air or by cable) are contacted by phone and interviewed regarding their listening habits. Each person is questioned about the signal sources which that individual listened to during the previous, say, twenty-four hours. However, this technique is suspect because it is subject to recall errors as well as possible bias introduced by the interviewer. If a specific signal source is mentioned to the person being interviewed, the suggestion may elicit a positive response even when tuning to that particular signal source actually did not occur. Another technique involves keeping diaries by persons agreeing to act as test subjects. Diary entries are to be made throughout the day to keep track of what signal sources are being listened to. The diaries are collected periodically and analyzed. However, this approach is prone to inaccuracies because the test subjects may fail to make entries due to forgetfulness or laziness. Thus, it can be readily seen that the recall-dependent approach first described above is unsatisfactory because people may not accurately remember what they listened to at any particular time and, also, because of the potential problem of suggestive bias. The diary-based approach is likewise unsatisfactory because people may not cooperate and be as meticulous in making diary entries as required to obtain the desired record-keeping accuracy.
It is also known to utilize a survey signal transmitted in combination with a programming signal for producing survey signals in the audible range. As disclosed in U.S. Pat. No. 4,718,106, the periodically transmitted survey signal is detected and reproduced by a speaker in the form of a code. The code is detected by a device worn by the individual being monitored, and data on the incidence of occurrence and/or the time of occurrence are stored and analyzed.
The tendency to reproduce the transmitted survey signal human audibly is a drawback of this technique because it can tend to disturb the listener. In order to provide meaningful results, an interval of preferably no more than ten minutes should elapse between survey signals. However, this can cause a chopping of, for example, a musical program at an inappropriate point, and some people can become annoyed just by virtue of this code being repeatedly reproduced audibly. Consequently, it is preferable to avoid use of a human audible survey signal. However, government regulations in some countries may require that signals for commercial radios, for example, must be limited to the human audible range. In fact, even though speakers which are now available can reproduce frequencies beyond the audible range of a human being, nevertheless the usable transmission frequencies permitted by government regulations are limited to such audible range because of the need for compatibility with older, lower quality speakers. Thus, there exists a conflict between the respective requirements at the transmission end and the receiving end. At the transmission end, there is the need to transmit a survey signal in the human audible frequency range, while at the receiving end it is preferable to reproduce the survey signal outside the human audible range.
Along with monitoring the signal source to which an individual is tuned, it is also useful to determine the length of time during which the individual remained tuned to such signal source. In U.S. Pat. No. 4,718,106 it is contemplated that, for example, the time of day is stored each time a coded survey signal is detected. If the signal is reproduced, say, every ten minutes, then each hour six time signals will need to be stored in memory. Since the memory must be capable of storing data collected over a reasonably long period of time, such as one month, it is readily apparent that a high capacity memory device would be required.